On 06/17/15 14:58, Michael S. Tsirkin wrote: > On Wed, Jun 17, 2015 at 01:34:33PM -0400, Don Slutz wrote: >> On 06/17/15 13:25, Paolo Bonzini wrote: >>> >>> >>> On 17/06/2015 19:14, Paolo Bonzini wrote: >>>> >>>> >>>> On 17/06/2015 19:03, Don Slutz wrote: >>>>> On 06/17/15 12:29, Michael S. Tsirkin wrote: >>>>>> On Wed, Jun 17, 2015 at 06:17:19PM +0200, Paolo Bonzini wrote: >>>>>>> >>>>>>> >>>>>>> On 17/06/2015 16:29, Michael S. Tsirkin wrote: >>>>>>>> On Wed, Jun 17, 2015 at 04:27:13PM +0200, Paolo Bonzini wrote: >>>>>>>>> >>>>>>>>> >>>>>>>>> On 17/06/2015 16:18, Michael S. Tsirkin wrote: >>>>>>>>>>>> Yes, that's what was done for parallel and pcspk as well. There's >>>>>>>>>>>> no >>>>>>>>>>>> infrastructure to avoid it. >>>>>>>>>>>> >>>>>>>>>>>> Paolo >>>>>>>>>> How do you mean? We have multiple ways to keep devices >>>>>>>>>> compatible with old versions. >>>>>>>>>> Set a new property to skip the extra stuff. >>>>>>>>> >>>>>>>>> Not if the device didn't have a vmstate at all, unfortunately. >>>>>>>> >>>>>>>> Skip creating the device completely for old machine types. >>>>>>> >>>>>>> Which device? The vmstate is tied to the same device that has always >>>>>>> been created. >>>>>> >>>>>> Just disable the new functionality. Make it behave in >>>>>> a compatible way. >>>>>> >>>>>>> we enable this thing by default (why do we?) >>>>>> >>>>>> Sigh. There is a very simple way to add a device in qemu: let user >>>>>> request it with -device. If one does this, one gets to maintain the >>>>>> resulting mess without bothering with pc maintainers in any way. >>>>>> >>>>>> But of course, everyone implementing a new feature feels it's such a >>>>>> great thing, and completel zero risk, it must be part of the default >>>>>> machine. Guess what, one then gets to bother with versioning from day 0. >>>>>> >>>>>>>>>> this seems like a big deal ... >>>>>>>>> >>>>>>>>> The PC speaker device is also enabled by default. >>>>>>>> >>>>>>>> This is historical, isn't it? >>>>>>> >>>>>>> Yes, but it has broken 2.3->2.2 migration. >>>>>>> >>>>>>> Let's just stop fighting windmills. >>>>>>> >>>>>>> Paolo >>>>>> >>>>>> I don't see what you are saying. Suddenly guest visible >>>>>> changes within a machine type are ok? >>>>>> >>>>>> So we have a bug, need to fix it, preferably before piling up >>>>>> more features. The best way imho is for 2.4 to avoid >>>>>> this device unless requested explicitly. >>>>>> >>>>> >>>>> My take on this is that Michael would like me to have a vmport_rpc=on >>>>> option, just like vmport=on (which already exists). With a default of >>>>> off. >>>> >>>> It wouldn't be enough, because dc->vmsd would be non-NULL anyway. >>>> >>>> (But yes, that option would be a good thing anyway). >>> >>> Even better would be to have a "-global vmport.rpc=no" option. It would >>> be simpler to disable it in existing machine types. >>> >> >> Either way I can avoid the device creation... Unless I hear otherwise I >> will go the global way. Since the default would be no, should I also >> make the default =yes for the 2.4 pc? >> >> -Don Slutz >> >> -Don Slutz > > Can you use -device vmport_rpc, and avoid adding code to the default pc? >
I have made this change (drop default creation, require -device to use) and so far testing looks good. -Don Slutz >>> Paolo >>>